File: ecdb.py

package info (click to toggle)
sagemath-database-cremona-elliptic-curves 20221013-1
  • links: PTS, VCS
  • area: main
  • in suites: bookworm, forky, sid, trixie
  • size: 5,252,172 kB
  • sloc: python: 3,515; makefile: 83; sh: 28
file content (987 lines) | stat: -rw-r--r-- 41,555 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
import os
import sys
sys.path.insert(0, '/home/jec/ecdata/scripts/')

from sage.all import (EllipticCurve, Integer, ZZ, Set, factorial,
                      mwrank_get_precision, mwrank_set_precision, srange, prod, copy, gcd)
from magma import get_magma

from red_gens import reduce_tgens, reduce_gens
from trace_hash import TraceHashClass

from files import (parse_line_label_cols, parse_curvedata_line,
                   parse_allgens_line_simple, parse_extra_gens_line, write_datafiles,
                   make_paricurves)

from codec import (parse_int_list, parse_int_list_list, point_to_weighted_proj,
                   weighted_proj_to_affine_point, split_galois_image_code,
                   curve_from_inv_string, shortstr, liststr, matstr, shortstrlist,
                   point_to_proj, mat_to_list_list)

from twoadic import get_2adic_data
from galrep import get_galrep_data
from intpts import get_integral_points
from aplist import my_aplist
from moddeg import get_modular_degree

from ec_utils import (mwrank_saturation_precision,
                      mwrank_saturation_maxprime, get_gens,
                      map_points)


from sage.databases.cremona import parse_cremona_label, class_to_int, cremona_letter_code

HOME = os.getenv("HOME")
sys.path.append(os.path.join(HOME, 'ecdata', 'scripts'))

FR_values = srange(1, 20) + [21, 25, 37, 43, 67, 163] # possible values of Faltings ratio

modular_degree_bound = 1000000 # do not compute modular degree if conductor greater than this

# Given a filename like curves.000-999, read the data in the file,
# compute the isogeny class for each curve, and output (1)
# allcurves.000-999, (2) allisog.000-999 (with the same suffix).  Also
# compute the bsd data for each curve and output (3) allbsd.000-999,
# (4) allgens.000-999 (with the same suffix), (5) degphi.000-999, (6)
# intpts.000-999, (7) alldegphi.000-999

# Version to compute gens & torsion gens too

def make_datafiles(infilename, mode='w', verbose=False, prefix="t"):
    infile = open(infilename)
    _, suf = infilename.split(".")
    allcurvefile = open(prefix+"allcurves."+suf, mode=mode)
    allisogfile = open(prefix+"allisog."+suf, mode=mode)
    allbsdfile = open(prefix+"allbsd."+suf, mode=mode)
    allgensfile = open(prefix+"allgens."+suf, mode=mode)
    degphifile = open(prefix+"degphi."+suf, mode=mode)
    alldegphifile = open(prefix+"alldegphi."+suf, mode=mode)
    apfile = open(prefix+"aplist."+suf, mode=mode)
    intptsfile = open(prefix+"intpts."+suf, mode=mode)
    for L in infile.readlines():
        if verbose:
            print("="*72)
        N, cl, _, ainvs, r, _, _ = L.split()
        E = EllipticCurve(parse_int_list(ainvs))
        r = int(r)
        # Compute the isogeny class
        Cl = E.isogeny_class()
        Elist = Cl.curves
        mat = Cl.matrix()
        maps = Cl.isogenies()
        ncurves = len(Elist)
        print("class {} (rank {}) has {} curve(s)".format(N+cl, r, ncurves))
        line = ' '.join([str(N), cl, str(1), ainvs, shortstrlist(Elist), matstr(mat)])
        allisogfile.write(line+'\n')
        if verbose:
            print("allisogfile: {}".format(line))
        # compute BSD data for each curve
        torgroups = [F.torsion_subgroup() for F in Elist]
        torlist = [G.order() for G in torgroups]
        torstruct = [list(G.invariants()) for G in torgroups]
        torgens = [[P.element() for P in G.gens()] for G in torgroups]
        cplist = [F.tamagawa_product() for F in Elist]
        omlist = [F.real_components()*F.period_lattice().real_period() for F in Elist]

        Lr1 = E.pari_curve().ellanalyticrank()[1].sage() / factorial(r)
        if r == 0:
            genlist = [[] for F in Elist]
            reglist = [1 for F in Elist]
        else:
            Plist = get_gens(E, r, verbose)
            # Saturate these points
            prec0 = mwrank_get_precision()
            mwrank_set_precision(mwrank_saturation_precision)
            if verbose:
                print("gens (before saturation) = {}".format(Plist))
            Plist, ind, _ = Elist[0].saturation(Plist, max_prime=-1)
            if verbose:
                print("gens (after saturation) = {}".format(Plist))
                if ind>1:
                    print("index gain {}".format(ind))
            genlist = map_points(maps, Plist)
            if verbose:
                print("genlist (before reduction) = {}".format(genlist))
            genlist = [Elist[i].lll_reduce(genlist[i])[0] for i in range(ncurves)]
            mwrank_set_precision(prec0)
            if verbose:
                print("genlist  (after reduction)= {}".format(genlist))
            reglist = [Elist[i].regulator_of_points(genlist[i]) for i in range(ncurves)]
        shalist = [Lr1*torlist[i]**2/(cplist[i]*omlist[i]*reglist[i]) for i in range(ncurves)]
        squares = [n*n for n in srange(1, 100)]
        for i, s in enumerate(shalist):
            if round(s) not in squares:
                print("bad sha value %s for %s" % (s, str(N)+cl+str(i+1)))
                print("Lr1 = %s" % Lr1)
                print("#t  = %s" % torlist[i])
                print("cp  = %s" % cplist[i])
                print("om  = %s" % omlist[i])
                print("reg = %s" % reglist[i])
                return Elist[i]
        if verbose:
            print("shalist = {}".format(shalist))

        # compute modular degrees
        # degphilist = [e.modular_degree(algorithm='magma') for e in Elist]
        # degphi = degphilist[0]
        # NB The correctness of the following relies on E being optimal!
        try:
            degphi = E.modular_degree(algorithm='magma')
        except RuntimeError:
            degphi = ZZ(0)
        degphilist1 = [degphi*mat[0, j] for j in range(ncurves)]
        degphilist = degphilist1
        if verbose:
            print("degphilist = {}".format(degphilist))

        # compute aplist for optimal curve only
        aplist = my_aplist(E)
        if verbose:
            print("aplist = {}".format(aplist))

        # Compute integral points (x-coordinates)
        intpts = [get_integral_points(Elist[i], genlist[i]) for i in range(ncurves)]
        if verbose:
            for i, Ei, xs in zip(range(ncurves), Elist, intpts):
                print("{}{}{} = {}: intpts = {}".format(N, cl, (i+1), Ei.ainvs(), xs))

        # Output data for optimal curves

        # aplist
        line = ' '.join([N, cl, aplist])
        if verbose:
            print("aplist: {}".format(line))
        apfile.write(line+'\n')

        # degphi
        if degphi:
            line = ' '.join([N, cl, '1', str(degphi), str(Set(degphi.prime_factors())).replace(' ', ''), shortstr(E)])
        else:
            line = ' '.join([N, cl, '1', str(degphi), str(Set([]).replace(' ', '')), shortstr(E)])
        if verbose:
            print("degphifile: {}".format(line))
        degphifile.write(line+'\n')

        # Output data for each curve
        for i in range(ncurves):
            # allcurves
            line = ' '.join([N, cl, str(i+1), shortstr(Elist[i]), str(r), str(torlist[i])])
            allcurvefile.write(line+'\n')
            if verbose:
                print("allcurvefile: {}".format(line))

            # allbsd
            line = ' '.join([N, cl, str(i+1), shortstr(Elist[i]), str(r), str(torlist[i]), str(cplist[i]), str(omlist[i]), str(Lr1), str(reglist[i]), str(shalist[i])])
            allbsdfile.write(line+'\n')
            if verbose:
                print("allbsdfile: {}".format(line))

            # allgens (including torsion gens, listed last)
            line = ' '.join([str(N), cl, str(i+1), shortstr(Elist[i]), str(r)]
                            + [liststr(torstruct[i])]
                            + [point_to_proj(P) for P in genlist[i]]
                            + [point_to_proj(P) for P in torgens[i]]
                           )
            allgensfile.write(line+'\n')
            if verbose:
                print("allgensfile: {}".format(line))
            # intpts
            line = ''.join([str(N), cl, str(i+1)]) + ' ' + shortstr(Elist[i]) + ' ' + liststr(intpts[i])
            intptsfile.write(line+'\n')
            if verbose:
                print("intptsfile: {}".format(line))

            # alldegphi
            line = ' '.join([str(N), cl, str(i+1), shortstr(Elist[i]), liststr(degphilist[i])])
            alldegphifile.write(line+'\n')
            if verbose:
                print("alldegphifile: {}".format(line))

    infile.close()
    allbsdfile.close()
    allgensfile.close()
    allcurvefile.close()
    allisogfile.close()
    degphifile.close()
    intptsfile.close()
    apfile.close()


# Create manin constant files from allcurves files:
#
# NB We assume that curve 1 in each class is optimal with constant 1
#
# Use this on a range where we have established that the optimal curve
# is #1.  Otherwise the C++ program h1pperiods outputs a file
# opt_man.<range> which includes what can be deduced about optimality
# (without a full check) and also outputs Manin constants conditional
# on the optimal curve being #1.

# The infilename should be an allcurves file, e.g. run in an ecdata
# directory and use infilename=allcurves/allcurves.250000-259999.  The
# part after the "." (here 250000-259999) will be used as suffix to
# the output file.

# Some classes may be output as "special": two curves in the class
# linked by a 2-isogeny with the first lattice having positive
# discriminant and the second Manin constant=2.  In several cases this
# has been an indication that the wrong curve has been tagged as
# optimal.

def make_manin(infilename, mode='w', verbose=False, prefix=""):
    infile = open(infilename)
    _, suf = infilename.split(".")
    allmaninfile = open(prefix+"opt_man."+suf, mode=mode)
    allisogfile = open("allisog/allisog."+suf)
    last_class = ""
    manins = []
    area0 = 0    # else pyflakes objects
    degrees = [] # else pyflakes objects
    for L in infile.readlines():
        N, cl, num, ainvs, _ = L.split(' ', 4)
        this_class = N+cl
        num = int(num)
        E = EllipticCurve(parse_int_list(ainvs))
        lattice1 = None
        if this_class == last_class:
            deg = degrees[int(num)-1]
            #assert ideg == deg
            area = E.period_lattice().complex_area()
            mc = round((deg*area/area0).sqrt())
            # print("{}{}{}: ".format(N, cl, num))
            # print("degree =     {}".format(deg))
            # print("area   =     {}".format(area))
            # print("area ratio = {}".format(area/area0))
            # print("c^2 =        {}".format(deg*area/area0))
            manins.append(mc)
            if num == 3:
                lattice1 = None
            elif not (num == 2 and deg == 2 and mc == 2):
                lattice1 = None
            if lattice1:
                print("Class {} is special".format(lattice1))
        else: # start a new class
            if manins and verbose: # else we're at the start
                print("class {} has Manin constants {}".format(last_class, manins))
            isogmat = allisogfile.readline().split()[-1]
            isogmat = parse_int_list_list(isogmat)
            degrees = isogmat[0]
            if verbose:
                print("class {}".format(this_class))
                print("isogmat: {}".format(isogmat))
                print("degrees: {}".format(isogmat))
            area0 = E.period_lattice().complex_area()
            manins = [1]
            if num == 1  and len(isogmat) == 2 and isogmat[0][1] == 2 and E.discriminant() > 0:
                lattice1 = this_class
            last_class = this_class

        # construct output line for this curve
        #
        # SPECIAL CASE 990h
        #
        if N == 90 and cl == 'h':
            opt = str(int(num == 3))
            manins = [1, 1, 1, 1]
        else:
            opt = str(int(num == 1))
        mc = str(manins[-1])
        line = ' '.join([str(N), cl, str(num), shortstr(E), opt, mc])

        allmaninfile.write(line+'\n')
        if verbose:
            print("allmaninfile: {}".format(line))
        # if int(N) > 100:
        #     break
    if manins and verbose: # else we're at the start
        # This output line is the only reason we keep the list of all m.c.s
        print("class {} has Manin constants {}".format(last_class, manins))
    infile.close()
    allmaninfile.close()

def make_opt_input(N):
    """Parse a file in optimality/ and produce an input file for
    runh1firstx1 which runs h1first1.

    Find lines containing "possible optimal curves", extract the
    isogeny class label, convert iso class code to a number, and
    output.  One line is output for each level N, of the form

    N i_1 i_2 ... i_k 0

    where N is the level and i_1,...,i_k are the numbers (from 1) of
    the newforms / isogeny classes where we do not know which curve is
    optimal.

    For example the input lines starting with 250010 are

    250010b: c=1; optimal curve is E1
    250010d: c=1; 3 possible optimal curves: E1 E2 E3

    and produce the output line

    250010 2 4 0

    while the lines starting with 250020 are

    250020b: c=1; optimal curve is E1
    250020d: c=1; optimal curve is E1

    and produce no output
    """
    outfile = "optx.{0:02d}".format(N)
    o = open(outfile, 'w')
    n = 0
    lastN = 0
    for L in open("optimality/optimality.{0:02d}".format(N)):
        if "possible" in L:
            N, c, _ = parse_cremona_label(L.split()[0][:-1])
            if N == lastN:
                o.write(" {}".format(1+class_to_int(c)))
            else:
                if lastN != 0:
                    o.write(" 0\n")
                o.write("{} {}".format(N, 1+class_to_int(c)))
                lastN = N
                n += 1
    o.write(" 0\n")
    n += 1
    o.close()
    print("wrote {} lines to {}".format(n, outfile))

def check_sagedb(N1, N2, a4a6bound=100):
    """Sanity check that Sage's elliptic curve database contains all
    curves [a1, a2, a3, a4, a6] with a1, a3 in [0, 1], a2 in [-1, 0, 1], a4,a6
    in [-100..100] and conductor in the given range.

    Borrowed from Bill Allombert (who found a missing curve of
    conductor 406598 this way).
    """
    from sage.all import CremonaDatabase
    CDB = CremonaDatabase()
    def CDB_curves(N):
        return [c[0] for c in CDB.allcurves(N).values()]
    Nrange = srange(N1, N2+1)
    ncurves = 12*(2*a4a6bound+1)**2
    print("Testing {} curves".format(ncurves))
    n = 0
    for a1 in range(2):
        for a2 in range(-1, 2):
            for a3 in range(2):
                for a4 in range(-a4a6bound, a4a6bound+1):
                    for a6 in range(-a4a6bound, a4a6bound+1):
                        ai = [a1, a2, a3, a4, a6]
                        n += 1
                        if n%1000 == 0:
                            print("test #{}/{}".format(n, ncurves))
                        try:
                            E = EllipticCurve(ai).minimal_model()
                        except ArithmeticError: #singular curve
                            continue
                        N = E.conductor()
                        if N not in Nrange:
                            continue
                        if list(E.ainvs()) not in CDB_curves(N):
                            print("Missing N={}, ai={}".format(N, ai))

# From here on, functions to process a set of "raw" curves, to create
# all the data files needed for LMFDB upload

# Assume that in directory CURVE_DIR there is a file 'curves.NN'
# containing one curve per line (valid formats: as in
# process_raw_curves() below).  For correct isogeny class labelling it
# is necessary that for every conductor present, at least one curve
# from each isogeny class is present.  The input list need not be
# closed under isogeny: if it is not then the 'allcurves.NN' file
# will contain more curves than the input file, otherwise they will
# contain the same curves. The file suffix 'NN' can be anything
# (nonempty) to identify the dataset, e.g. it could be a single
# conductor, or a range of conductors.
#
# Step 1: sort and label.  Run sage from ecdata/scripts:
#
# sage: %runfile ecdb.py
# sage: NN = '2357' # (say)
# sage: curves_file = 'curves.{}'.format(NN)
# sage: allcurves_file = 'allcurves.{}'.format(NN)
# sage: process_raw_curves(curves_file, allcurves_file, base_dir=CURVE_DIR)
#
# If you add split_by_N=True to process_raw_curves() then instead of
# one allcurves file you will get one per conductor N, each of the
# form allcurves.<N>.  This can be useful for using parallel
# processing in the next step, after which you have to concatenate all
# the output files.  In this case allcurves_file will be ignored, and
# a file <curves_file>.conductors will also be written containing the
# distinct conductors in the input file.
#
#
# Step 2: compute most of the data, run also in ecdb/scripts after reading ecdb.py:
#
# sage: read_write_data(allcurves_file, CURVE_DIR)
#
# We now have files CURVE_DIR/F/F.NN for F in ['curvedata', 'classdata', 'intpts', 'alldegphi'].
#
# Step 2a: if we have processed allcurves.RANGE with more than one
#  conductor, then for each conductor N we will have files F/F.N for F
#  in ['curvedata', 'classdata', 'intpts', 'alldegphi'].  We now check
#  that these are complete and merge them into just 4 files F/F.RANGE.
#
# (o) RANGE=...
#     CURVE_DIR=...
#     cd ${CURVE_DIR}
#
# (i) extract conductors
# awk '{print $1;}' allcurves/allcurves.${RANGE} | sort -n | uniq > conductors.RANGE
#
# (ii) check files are complete (this is quick)

# for F in curvedata classdata intpts alldegphi; do echo $F; for N in `cat conductors.${RANGE}`; do if ! [ -e ${CURVE_DIR}/${F}/${F}.${N} ]; then echo ${F}"."${N}" missing"; fi; done; done;
#
# (iii) merge files (this takes ages for ~200000 conductors)
# Here we don't bother with alldegphi files as they will be empty for large N.

# for F in curvedata classdata intpts; do echo $F; for N in `cat conductors.${RANGE}`; do cat ${CURVE_DIR}/${F}/${F}.${N} >> ${CURVE_DIR}/${F}/${F}.${RANGE}; done; done
#
# Better way if you can find all the files easily with a wildcard:
# e.g. for all conductors in range 1e7-1e8:
# echo curvedata.???????? | xargs cat > curvedata.${RANGE}
#
# or
# pushd curvedata
# echo $(for N in $(cat ../conductors.${RANGE}); do echo curvedata\.${N}; done) | xargs cat > curvedata.${RANGE}
# popd
#
# Step 3: run magma scripts to compute galrep and 2adic data. From ecdata/scripts:
# ./make_galrep.sh NN CURVE_DIR
#
# This uses CURVE_DIR/allcurves/allcurves.NN and
# creates   CURVE_DIR/ft/ft.NN for ft in ['galrep', '2adic'].
#
# Step 4: create 6 files in UPLOAD_DIR (default ${HOME}/ecq-upload):
#
# ec_curvedata.NN, ec_classdata.NN, ec_localdata.NN,
# ec_mwbsd.NN, ec_galrep.NN, ec_2adic.NN
#
# In ecdata/scripts run sage:
# sage: %runfile files.py
# sage: data = read_data(CURVE_DIR, file_types=main_file_types, ranges=[NN])
# sage: make_all_upload_files(data, tables = main_tables, NN=NN)
#
# ready for upload to LMFDB's db.ec_curvedata (etc) using
#
# db.ec_curvedata.update_from_file()


def process_raw_curves(infilename, outfilename, base_dir='.', split_by_N=False, verbose=1):
    """File infilename should contain one curve per line, with
    a-ainvariants or c-invariants as a list (with no internal spaces),
    optionally preceded by the conductor.

    Sample lines (all defining the same curve):

    [0,-1,1,-10,-20]
    11 [0,-1,1,-10,-20]
    [496,20008]
    11 [496,20008]

    We do not assume minimal or reduced models, and if the conductor
    is given it is checked for consistency.

    For each input curve we check whether an isogenous curve has
    already been processed; if not, we compute its isogeny class and
    process all of them.

    OUTPUT: one line per curve, sorted by conductor and then by
    isogeny class.  Each line has the format

    N class_code class_size number lmfdb_number ainvs

    where "number" is the index of the curve in its class (counting
    from 1), sorted by Faltings heights, with the lex. order of
    a-invariants as a tie-breaker.

    If split_by_N is False, the output will be all in one file.  If
    True, there will be one output file per conductor N, whose name is
    allcurves file with suffix ".{N}".

    """
    # allcurves will have conductors as keys, values lists of lists of
    # ainvs, subdivided by isogeny class
    allcurves = {}
    ncurves = ncurves_complete = 0

    with open(os.path.join(base_dir, infilename)) as infile:
        for L in infile:
            data = L.split()
            assert len(data) in [1, 2]
            invs = data[-1]
            E = curve_from_inv_string(invs)
            ainvs = E.ainvs()
            N = E.conductor()
            if len(data) == 2:
                N1 = ZZ(data[0])
                if N1 != N:
                    raise ValueError("curve with invariants {} has conductor {}, not {} as input".format(invs, N, N1))
            ncurves += 1
            if ncurves%1000 == 0 and verbose:
                print("{} curves read from {} so far...".format(ncurves, infilename))

            # now we have the curve E, check if we have seen it (or an isogenous curve) before
            repeat = False

            if N in allcurves:
                if any(ainvs in c for c in allcurves[N]):
                    repeat = True
                    if verbose > 1:
                        print("Skipping {} of conductor {}: repeat".format(ainvs, N))
                else:
                    if verbose > 1:
                        print("New curve {} of conductor {}".format(ainvs, N))
            else:
                allcurves[N] = []
            if repeat:
                continue # to next input line
            newcurves = [E2.ainvs() for E2 in E.isogeny_class().curves]
            ncurves_complete += len(newcurves)
            allcurves[N].append(newcurves)

    print("{} curves read from {}".format(ncurves, infilename))
    if ncurves != ncurves_complete:
        print("input curves not closed under isogeny! completed list contains {} curves".format(ncurves_complete))

    conductors = list(allcurves.keys())
    conductors.sort()

    for N in conductors:
        C = allcurves[N]
        ncl = len(C)
        nc = sum([len(cl) for cl in C])
        print("Conductor {}:\t{} isogeny classes,\t{} curves".format(N, ncl, nc))

    # Now we work out labels and output a new file

    def curve_key_LMFDB(E): # lex order of ainvs
        return E.ainvs()

    def curve_key_Faltings(E): # Faltings height, with tie-break
        return [-E.period_lattice().complex_area(), E.ainvs()]

    def output_one_conductor(N, allcurves_N, outfile):
        nNcl = nNcu = 0
        sN = str(N)
        # construct the curves from their ainvs:
        CC = [[EllipticCurve(ai) for ai in cl] for cl in allcurves_N]
        nap = 100
        ok = False
        while not ok:
            aplists = dict([(cl[0], cl[0].aplist(100, python_ints=True)) for cl in CC])
            aps = list(aplists.values())
            ok = (len(list(Set(aps))) == len(aps))
            if not ok:
                print("{} ap not enough for conductor {}, increasing...".format(nap, N))
            nap += 100
        # sort the isogeny classes:
        CC.sort(key=lambda cl: aplists[cl[0]])
        for ncl, cl in enumerate(CC):
            nNcl += 1
            class_code = cremona_letter_code(ncl)
            # sort the curves in two ways (LMFDB, Faltings)
            cl.sort(key=curve_key_LMFDB)
            cl_Faltings = copy(cl)
            cl_Faltings.sort(key=curve_key_Faltings)
            class_size = len(cl)
            for nE_F, E in enumerate(cl_Faltings):
                nNcu += 1
                ainvs = shortstr(E)
                nE_L = cl.index(E)
                line = " ".join([sN, class_code, str(class_size), str(nE_F+1), str(nE_L+1), ainvs])
                #print(line)
                outfile.write(line+"\n")
        return nNcu, nNcl

    if split_by_N:
        with open(os.path.join(base_dir, infilename+".conductors"), 'w') as Nfile:
            for N in conductors:
                Nfile.write("{}\n".format(N))
                outfilename_N = "allcurves/allcurves.{}".format(N)
                with open(os.path.join(base_dir, outfilename_N), 'w') as outfile:
                    nNcu, nNcl = output_one_conductor(N, allcurves[N], outfile)
                    print("N={}: {} curves in {} classes output to {}".format(N, nNcu, nNcl, outfilename_N))

    else:
        with open(os.path.join(base_dir, outfilename), 'w') as outfile:
            for N in conductors:
                nNcu, nNcl = output_one_conductor(N, allcurves[N], outfile)
                print("N={}: {} curves in {} classes output to {}".format(N, nNcu, nNcl, outfilename))

def make_new_data(infilename, base_dir, Nmin=None, Nmax=None, PRECISION=128, verbose=1,
                  allgensfilename=None, oldcurvedatafile=None, extragensfilename=None):
    alldata = {}
    nc = 0
    labels_by_conductor = {}
    tempdir = os.path.join(base_dir, "temp")

    with open(os.path.join(base_dir, "allcurves", infilename)) as infile:
        for L in infile:
            sN, isoclass, class_size, number, lmfdb_number, ainvs = L.split()
            N = ZZ(sN)
            if Nmin and N < Nmin:
                continue
            if Nmax and N > Nmax:
                continue
            nc += 1
            iso = ''.join([sN, isoclass])
            label = ''.join([iso, number])
            lmfdb_isoclass = isoclass
            lmfdb_iso = '.'.join([sN, isoclass])
            lmfdb_label = ''.join([lmfdb_iso, lmfdb_number])
            iso_nlabel = class_to_int(isoclass)
            number = int(number)
            lmfdb_number = int(lmfdb_number)
            class_size = int(class_size)
            ainvs = parse_int_list(ainvs)
            bad_p = N.prime_factors() # will be sorted

            record = {
                'label': label,
                'isoclass': isoclass,
                'iso': iso,
                'number': number,
                'iso_nlabel': iso_nlabel,
                'lmfdb_number': lmfdb_number,
                'lmfdb_isoclass': lmfdb_isoclass,
                'lmfdb_iso': lmfdb_iso,
                'lmfdb_label':lmfdb_label,
                'faltings_index': number,
                'class_size': class_size,
                'ainvs': ainvs,
                'conductor': N,
                'bad_primes': bad_p,
                'num_bad_primes': len(bad_p),
                }
            if label in alldata:
                raise RuntimeError("duplicate label {} in {}".format(label, infilename))
            alldata[label] = record
            if N in labels_by_conductor:
                labels_by_conductor[N].append(label)
            else:
                labels_by_conductor[N] = [label]

    allN = list(labels_by_conductor.keys())
    allN.sort()
    firstN = min(allN)
    lastN = max(allN)
    if verbose:
        print("{} curves read from {}, with {} distinct conductors from {} to {}".format(nc, infilename, len(labels_by_conductor), firstN, lastN))

    gens_dict = {}
    if allgensfilename:
        print("Reading from {}".format(allgensfilename))
        n = 0
        with open(os.path.join(base_dir, allgensfilename)) as allgensfile:
            for L in allgensfile:
                if Nmin or Nmax:
                    label, record = parse_line_label_cols(L)
                    N = record['conductor']
                    if Nmin and N < Nmin:
                        continue
                    if Nmax and N > Nmax:
                        continue
                n += 1
                label, record = parse_allgens_line_simple(L)
                gens_dict[tuple(record['ainvs'])] = record['gens']
                if n%10000 == 0:
                    print("Read {} curves from {}".format(n, allgensfilename))

    if oldcurvedatafile:
        print("Reading from {}".format(oldcurvedatafile))
        n = 0
        with open(os.path.join(base_dir, "curvedata", oldcurvedatafile)) as oldfile:
            for L in oldfile:
                label, record = parse_curvedata_line(L)
                N = int(record['conductor'])
                if Nmin and N < Nmin:
                    continue
                if Nmax and N > Nmax:
                    continue
                n += 1
                gens = [weighted_proj_to_affine_point(P) for P in record['gens']]
                gens_dict[tuple(record['ainvs'])] = gens
                if n%10000 == 0:
                    print("Read {} curves from {}".format(n, oldcurvedatafile))

    if extragensfilename:
        print("Reading from {}".format(extragensfilename))
        n = 0
        with open(os.path.join(base_dir, "allgens", extragensfilename)) as genfile:
            for L in genfile:
                N = ZZ(L.split()[0])
                if Nmin and N < Nmin:
                    continue
                if Nmax and N > Nmax:
                    continue
                N, ainvs, gens = parse_extra_gens_line(L)
                n += 1
                if gens:
                    gens_dict[ainvs] = gens
                if n%10000 == 0:
                    print("Read {} curves from {}".format(n, extragensfilename))

    N0 = 0
    for N, labels in labels_by_conductor.items():
        if N != N0 and N0:
            # extract from alldata the labels for conductor N0, and output it
            if verbose:
                print("Finished conductor {}".format(N0))
                print("Writing data files for conductor {}".format(N0))
            write_datafiles(dict([(lab, alldata[lab]) for lab in labels_by_conductor[N0]]),
                            N0, tempdir)
        N0 = N
        if verbose:
            print("Processing conductor {}".format(N))
        for label in labels:
            record = alldata[label]
            if verbose:
                print("Processing curve {}".format(label))
            iso = record['iso']
            number = record['number']
            first = (number == 1) # tags first curve in each isogeny class
            ncurves = record['class_size']
            if first:
                alllabels = [iso+str(k+1) for k in range(ncurves)]
                allcurves = [EllipticCurve(alldata[lab]['ainvs']) for lab in alllabels]
                E = allcurves[0]
            else:
                record1 = alldata[iso+'1']
                E = allcurves[number-1]
            assert N == E.conductor()
            if verbose > 1:
                print("E = {}".format(E.ainvs()))

            record['jinv'] = Ej = E.j_invariant()
            record['potential_good_reduction'] = (Ej.denominator() == 1)
            D = E.discriminant()
            record['absD'] = ZZ(D).abs()
            record['signD'] = int(D.sign())
            record['cm'] = int(E.cm_discriminant()) if E.has_cm() else 0

            if first:
                record['aplist'] = E.aplist(100, python_ints=True)
                record['anlist'] = E.anlist(20, python_ints=True)
                record['trace_hash'] = TraceHashClass(record['iso'], E)
            else:
                record['aplist'] = record1['aplist']
                record['anlist'] = record1['anlist']
                record['trace_hash'] = record1['trace_hash']

            if verbose > 1:
                print("aplist done")
            local_data = [{'p': int(ld.prime().gen()),
                           'ord_cond':int(ld.conductor_valuation()),
                           'ord_disc':int(ld.discriminant_valuation()),
                           'ord_den_j':int(max(0, -(Ej.valuation(ld.prime().gen())))),
                           'red':int(ld.bad_reduction_type()),
                           'rootno':int(E.root_number(ld.prime().gen())),
                           'kod':ld.kodaira_symbol()._pari_code(),
                           'cp':int(ld.tamagawa_number())}
                          for ld in E.local_data()]

            record['tamagawa_numbers'] = cps = [ld['cp'] for ld in local_data]
            record['kodaira_symbols'] = [ld['kod'] for ld in local_data]
            record['reduction_types'] = [ld['red'] for ld in local_data]
            record['root_numbers'] = [ld['rootno'] for ld in local_data]
            record['conductor_valuations'] = cv = [ld['ord_cond'] for ld in local_data]
            record['discriminant_valuations'] = [ld['ord_disc'] for ld in local_data]
            record['j_denominator_valuations'] = [ld['ord_den_j'] for ld in local_data]

            record['semistable'] = all([v == 1 for v in cv])
            record['tamagawa_product'] = tamprod = prod(cps)
            if verbose > 1:
                print("local data done")

            twoadic_data = get_2adic_data(E, get_magma())
            record.update(twoadic_data)
            if verbose > 1:
                print("2-adic data done")

            record['modp_images'] = image_codes = get_galrep_data(E, get_magma())
            record['nonmax_primes'] = pr = [int(split_galois_image_code(s)[0]) for s in image_codes]
            record['nonmax_rad'] = prod(pr)
            if verbose > 1:
                print("galrep data done")

            T = E.torsion_subgroup()
            tgens = [P.element() for P in T.gens()]
            tgens.sort(key=lambda P: P.order())
            tgens = reduce_tgens(tgens)
            tor_struct = [P.order() for P in tgens]
            record['torsion_generators'] = [point_to_weighted_proj(gen) for gen in tgens]
            record['torsion_structure'] = tor_struct
            record['torsion'] = torsion = prod(tor_struct)
            record['torsion_primes'] = [int(p) for p in Integer(torsion).support()]
            if verbose > 1:
                print("torsion done")

            if first: # else add later to avoid recomputing a.r.

                # Analytic rank and special L-value
                ar, sv = E.pari_curve().ellanalyticrank(precision=PRECISION)
                record['analytic_rank'] = ar = ar.sage()
                record['special_value'] = sv = sv.sage()/factorial(ar)

                # compute isogenies so we can map points from #1 to the rest:
                cl = E.isogeny_class(order=tuple(allcurves))
                record['isogeny_matrix'] = mat = mat_to_list_list(cl.matrix())
                record['class_deg'] = max(max(r) for r in mat)
                record['isogeny_degrees'] = mat[0]
                isogenies = cl.isogenies() # for mapping points later
                if N <= modular_degree_bound:
                    record['degree'] = degphi = get_modular_degree(E, label)
                    if verbose > 1:
                        print("degphi = {}".format(degphi))
                else:
                    record['degree'] = degphi = 0
                    if verbose > 1:
                        print("degphi not computed as conductor > {}".format(modular_degree_bound))
                record['degphilist'] = [degphi*mat[0][j] for j in range(ncurves)]
            else:

                record['analytic_rank'] = ar = record1['analytic_rank']
                record['special_value'] = sv = record1['special_value']
                record['class_deg'] = record1['class_deg']
                record['isogeny_degrees'] = record1['isogeny_matrix'][number-1]
                record['degree'] = record1['degphilist'][number-1]

            if verbose > 1:
                print("analytic rank done: {}".format(ar))

            gens_missing = False

            if ar == 0:
                record['gens'] = gens = []
                record['regulator'] = reg = 1
                record['ngens'] = 0
                record['heights'] = []
                record['rank'] = 0
                record['rank_bounds'] = [0, 0]

            else: # positive rank
                if first:
                    if verbose > 1:
                        print("{}: an.rk.={}, finding generators".format(label, ar))
                    #if 'gens' in record:
                    ainvs = tuple(record['ainvs'])
                    if ainvs in gens_dict:
                        gens = [E(P) for P in gens_dict[ainvs]]
                        if verbose > 1:
                            print("..already have gens {}, just saturating (p<{})...".format(gens, mwrank_saturation_maxprime))
                        gens, n, _ = E.saturation(gens, max_prime=mwrank_saturation_maxprime)
                        ngens = len(gens)
                        if ngens < ar:
                            if verbose:
                                print("Warning: a.r.={} but we only have {} gens".format(ar, ngens))
                        if verbose > 1:
                            if n and n > 1:
                                print("..saturation index was {}, new gens: {}".format(n, gens))
                            else:
                                print("..saturated already")
                    else:
                        gens = get_gens(E, ar, verbose) # this returns saturated points
                    ngens = len(gens)
                    if ngens < ar:
                        gens_missing = True
                        print("{}: analytic rank = {} but we only found {} generators".format(label, ar, ngens))
                    else:
                        if verbose > 1:
                            print("...done, generators {}".format(gens))
                    record['rank_bounds'] = [ngens, ar]
                    record['rank'] = None if gens_missing else ngens
                    # so the other curves in the class know their gens:
                    record['allgens'] = map_points(isogenies, gens, verbose) # returns saturated sets of gens
                else:
                    gens = record1['allgens'][number-1]
                    record['rank'] = record1['rank']
                    record['rank_bounds'] = record1['rank_bounds']
                    gens_missing = (record['rank'] is None)

                gens, tgens = reduce_gens(gens, tgens, False, label)
                record['gens'] = [point_to_weighted_proj(gen) for gen in gens]
                record['heights'] = heights = [P.height(precision=PRECISION) for P in gens]
                reg = None if gens_missing else E.regulator_of_points(gens, PRECISION)
                record['ngens'] = len(gens)
                record['regulator'] = reg
                if verbose > 1:
                    print("... finished reduction, gens are now {}, heights {}, reg={}".format(gens, heights, reg))

            L = E.period_lattice()
            record['real_period'] = om = L.omega(prec=PRECISION) # includes #R-components factor
            record['area'] = A = L.complex_area(prec=PRECISION)
            F_ht = -A.log()/2
            R = om.parent()
            record['faltings_height'] = F_ht
            record['stable_faltings_height'] = F_ht - R(gcd(D, E.c4()**3)).log()/12

            # Analytic Sha
            if gens_missing:
                if verbose:
                    print("Unable to compute analytic Sha since #gens < analytic rank")
                record['sha_an'] = None
                record['sha'] = None
            else:
                sha_an = sv*torsion**2 / (tamprod*reg*om)
                sha = sha_an.round()
                warn = "sha_an = {}, rounds to {}".format(sha_an, sha)
                assert sha > 0, warn
                assert sha.is_square(), warn
                assert ((sha-sha_an).abs() < 1e-10), warn
                record['sha_an'] = sha_an
                record['sha'] = int(sha)

            # Faltings ratio
            FR = 1 if first else (record1['area']/A).round()
            assert FR in FR_values, "F ratio = {}/{} = {}".format(record1['area'], A, FR)
            record['faltings_ratio'] = FR

            if verbose > 1:
                print(" -- getting integral points...")
            record['xcoord_integral_points'] = get_integral_points(E, gens)
            if verbose > 1:
                print(" ...done: {}".format(record['xcoord_integral_points']))


            Etw, Dtw = E.minimal_quadratic_twist()
            if Etw.conductor() == N:
                record['min_quad_twist_ainvs'] = record['ainvs']
                record['min_quad_twist_disc'] = 1
            else:
                record['min_quad_twist_ainvs'] = [int(a) for a in Etw.ainvs()]
                record['min_quad_twist_disc'] = int(Dtw)

            if verbose:
                print("Finished processing {}".format(label))
                if verbose > 1:
                    print("data for {}:\n{}".format(label, record))

    # don't forget to output last conductor
    if verbose:
        print("Finished conductor {}".format(N0))
        print("Writing data files for conductor {}".format(N0))
    write_datafiles(dict([(lab, alldata[lab]) for lab in labels_by_conductor[N0]]),
                    N0, tempdir)
    outfilename = infilename.replace("allcurves.", "")
    if Nmin or Nmax:
        outfilename = "{}.{}-{}".format(outfilename, firstN, lastN)
    if verbose:
        print("Finished all, writing data files for {}".format(outfilename))

    write_datafiles(alldata, outfilename, base_dir)

    return alldata

def read_write_data(infilename, base_dir, verbose=1):
    print("Reading from {}".format(infilename))
    N = ".".join(infilename.split(".")[1:])
    data = make_new_data(infilename, base_dir=base_dir, verbose=verbose)
    write_datafiles(data, N, base_dir)

################################################################################